ABSTRACT:Liver slice experiments were performed to determine the slice intrinsic clearance and to extrapolate this to the in vivo liver intrinsic clearance in a physiologically based pharmacokinetic (PBPK)-like approach. Precision-cut liver slices were incubated with different initial concentrations of tolbutamide, and the time series of parent and metabolite concentrations were measured in slice and incubation medium. A mathematical model was built that modeled the uptake of tolbutamide and its metabolism in the liver slice. In addition, binding of tolbutamide to cellular constituents and partition over the water and lipid phase were accounted for by the model. Model analysis imposed sampling of parent compound in slice and of metabolites pooled from slice and medium. The model was calibrated to the data, fitting the intrinsic clearance, the parent In the past decade, precision-cut liver slices have been successfully used as an in vitro model for biotransformation (Dogterom and Rothuizen, 1993;Ball et al., 1996;Hashemi et al., 1999), hepatotoxicity (Wormser et al., 1990;Miller et al., 1993), and enzyme induction studies (Lake et al., 1997) in a variety of animal species. Important determinants in this successful application of liver slices are undoubtedly the maintenance of normal tissue architecture, cell heterogeneity, and cell-cell communications within the livers' original tissue matrix.However, a drawback of liver slices often addressed in studies predicting rates of metabolism is their diffusional limitations (Ekins et al., 1995;Worboys et al., 1997). The outer cell layers of liver slices are directly exposed to incubation media; however, the inner cell layers are only exposed to the test compound when it has traveled through or around the outer cell layers. This phenomenon has been visualized using a fluorescent dye (Ekins et al., 1995). Diffusional limitations have often led to underevaluation of in vivo intrinsic clearance rates when liver slice data are normalized for hepatocellularity (Worboys et al., 1996a,b).These problems, however, can be counteracted when appropriate modeling of in vitro data is incorporated in the prediction strategy. In this respect, it is important to note that Worboys et al. (1997) assessed drug metabolism in rat liver slices by analyzing experimental data with a classical one-compartment model. However, the liver slice model consists of two physically different phases (i.e., slice and surrounding incubation medium), and hence, a two-compartment model seems more appropriate. In contrast to other liver slice models designed to predict rates of metabolism we developed a model that consists of a physicochemical part, a mathematical part, and an observable part. Our liver slice model takes into consideration processes of transport, partitioning, and elimination of drug and/or its metabolites, which leads to the identification of the ultimate parameter: the slice metabolic rate constant CL s . 1 The development of our rat liver slice model can be seen as a first step to improve t...